Detailed investigation at the site of CTRL‘s new international station in Stratford, east London, could mean that cost effective well pumping could be used in dealing with groundwater.

In designing a dewatering control system that relies on pumping, detailed information is needed on the ground conditions, on the water bearing strata (or aquifers) present and the hydraulic properties of each stratum. Engineers can then determine the optimum number, type, depth and spacing of wells as well as their pumping capacity.

The yield of individual wells is also considered at the design stage because costs are reduced if a smaller number of more efficient wells can be drilled.

At Stratford, Hampshire firm Foundation & Exploration Services was called on to carry out large scale pumping out tests from trial wells to determine permeability and storage coefficient. But the tests, for CTRL project manager Rail Link Engineering, also included trials of the unusual methods of acidisation and vacuum pumping.

Results of the investigations at Stratford identified a high water table and the ground conditions.

It was known that groundwater would have a major effect on tunnel construction because the tunnel level lies mainly within the Thanet Sand. Running tunnels rise up through the overlying strata to an open excavation area near ground level, and then descend to deeper levels once again.

To determine the best construction methods, engineers assessed the dewatering and pressure relief from wells in the terrace gravel, Thanet Sand and Chalk aquifers. Groundwater control would be needed in the terrace gravel aquifer and also to reduce the water pressure in the Thanet Sand to prevent the base of the excavation heaving up or rupturing from the pressures in this layer.

The Chalk and Thanet Sand aquifers, while having greatly differing hydraulic properties, were thought to be hydraulically connected. The Thanet Sand is a very dense, very fine-grained, often slightly silty formation, with a relatively low permeability. With high capillary forces around the grains, the water yield to individual wells can be low.

Rather than attempt to lower the water pressure in the Thanet Sand by pumping directly from this stratum, experience on other sites in similar ground conditions in London suggested that pumping from the underlying high permeability Chalk would be a more cost effective method. As the two formations are hydraulically connected, pumping from the Chalk produces an under-draining effect in the Sand and this can cover a wide area because of the broad cone of depression. A series of pumping tests were carried out to determine the feasibility of this method.

Pumping tests

Pumping tests measure the response at various points remote from the pumping well, called observation wells. Each pumping well and observation well is sealed into the aquifer using cement and bentonite grout to prevent any hydraulic connection along the borehole between each layer. Electric submersible borehole pumps remove the water. The response in each observation well is measured using an electronic water pressure transducer connected to a data logger. A pumping test in each aquifer is carried out separately, monitoring the response in all the other formations. Further valuable information can be obtained by monitoring the recovery of the aquifers.

A full scale recharge test, to pump the water in instead of out, was also carried out in the gravel well. If it is feasible to return the water to the aquifer, disposal costs could be greatly reduced.

Techniques to improve well yield

When test pumping, it is usually beneficial to be able to remove as much water as possible from the aquifer via the single well constructed. With a pronounced lowering, engineers can determine the effects of recharge boundaries (such as rivers) or barrier boundaries (such as the edge of the aquifer) at some distance from the well. And pumping a wider area gives generally more reliable data.

Steps are needed to ensure that the well, and the well/aquifer interface, are as efficient a water conducting medium as possible. A well usually comprises a slotted section of plastic or metal pipe in a borehole, with as large an open area as possible. This is surrounded by a granular filter medium which physically retains and supports the aquifer material while allowing the free passage of water to the open pipe. Conventional development involves agitating the granular filter medium and the aquifer itself using air surging, water jetting or mechanical disturbance to remove all traces of soils, drilling mud and debris from the borehole wall. In rock formations like chalk, an open unsupported borehole can be used if the stratum is stable.

However, at Stratford it was found that both the Thanet Sand and Chalk boreholes needed to be stimulated to improve their yield.

Thanet Sand borehole

Conventional development of the well produced an unusually low yield and so it was decided to seal off the well and produce a vacuum inside it; then to repeat the tests. While specialist vacuum pumping techniques are often used during full scale construction dewatering, this is not normally the case during tests because much less powerful equipment is used. FES therefore needed to devise a small scale vacuum technique, more suited to the test phase and using much smaller plant, but which would simulate full scale construction dewatering.

The vacuum increases the pressure differential between the aquifer and the well, helping to break down capillary forces, and encourages a greater inflow of water from the fine grained sands. Under 0.5 bar vacuum the well yield improved 40% compared with that when pumped conventionally.

Chalk borehole

After extensive conventional development, CCTV had shown that no traces of drilling debris or smeared putty chalk had remained on the borehole wall. Poor natural fissure development was suspected, but data from some of the observation wells showed a very high level of fissuring present nearby.

It was therefore decided to apply an acid injection to the well to improve its yield. The method is regularly used by the water supply industry but seldom adopted during ground investigation and construction dewatering projects.

Commercial grade hydrochloric acid was injected at three depths selected where fissure development was likely to be highest. Acid produces a strong chemical reaction with the calcium carbonate which dissolves and widens the fissures. It also produces carbon dioxide and if a sealed wellhead is used, then the pressure this gas generates can be used to force the acid deeper into the aquifer fissures. Any chalk slurry produced by drilling which has entered the fissures is also removed by the process. The acid is completely used up during the chemical reaction and has no lasting effect on groundwater quality.

Successful results

The acidisation technique improved the well yield from 1.25litre/s during air lifting to 9 litre/s with trial pumping. The tests were run for long periods to assess the effect on the water table.

From the test data a hydrogeological model for the site can be produced to assess the optimum tunnel alignment, construction method and dewatering schemes. In addition, the investigation provided valuable information on both vacuum dewatering and acidisation in improving well yield.

Related Files
Figure 1 – Acid Injection